A datacenter is generally a room containing a plurality of cabinet racks of computer systems, like computer servers, for example. For high end computer servers, each cabinet rack may comprise from two to sixty-four cell/memory units, which are commonly referred to as boards. Each board may have a unique identification (ID). A exemplary datacenter is described in the co-pending U.S. patent application Ser. No. 10/614,856, filed Jul. 9, 2003 and published Jan. 13, 2005, entitled “Inventory Management of Components” and assigned to the same assignee as the instant application, which application being incorporated by reference herein in its entirety.
Currently, the higher end computer servers have self-diagnostic capabilities for identifying a failed board. However, maintenance of the failed server remains a manual process. Typically, the failed server sends a failure notification to a system or service administrator which includes the ID of the failed board or unit. The administrator may or may not be local to the datacenter. If remote from the datacenter, the notification may be an email message, for example. For the most part, datacenters do not include a storage inventory of the different types of replacement boards. Accordingly, in such cases, upon receipt of a failure notification, the administrator knowing the ID of the failed board will order a replacement unit and have it shipped to the datacenter. A service technician will take the replacement unit to the failed server for manual replacement. The failed unit may have an indicator light for directing the technician to the failed unit.
Of course, for those servers that do not have self-diagnostics capabilities, the failed server will have to be identified by a service technician, for example. Then, diagnostics will be performed on the identified server to locate the failed board or boards. Thereafter, the manual procedure will remain the same or similar to that described above.
This manual procedure for maintaining the datacenter operational is rather onerous and time consuming, requiring one or more service technicians either on-duty or on-call. The lost operational time resulting from a failed server is also a consequential loss to the datacenter operator. Accordingly, an improvement to the manual operational maintenance of the datacenter is desired for at least the aforementioned reasons.